cybernetics is the science of control. Its name, appropriately suggested by the mathematician Norbert Wiener (1894–1964), is derived from the Greek for ‘steersman’, pointing to the essence of cybernetics as the study and design of devices for maintaining stability, or for homing in on a goal or target. Its central concept is
feedback. Since the ‘devices’ may be living or man-made, cybernetics bridges biology and engineering.
Stability of the human body is achieved by its static geometry and, very differently, by its dynamic control. A statue of a human being has to have a large base or it topples over. It falls when the centre of mass is vertically outside the base of the feet. Living people make continuous corrections to maintain themselves standing. Small deviations of
posture are signalled by sensory signals (
proprioception) from nerve fibres in the muscles and around the joint capsules of the ankles and legs, and by the
otoliths (the organs of
balance in the inner ear). Corrections of posture are the result of dynamic feedback from these senses, to maintain dynamic stability. When walking towards a target, such as the door of a room, deviations from the path are noted, mainly visually, and corrected from time to time during the movement, until the goal is reached. The key to this process is continuous correction of the output system by signals representing detected errors of the output, known as ‘negative feedback’. The same principle, often called
servo-control, is used in engineering, in order to maintain the stability of machinery and to seek and find goals, with many applications such as guided missiles and autopilots.
The principles of feedback apply to the body's regulation of temperature,
blood pressure, and so on. Though the principles are essentially the same as in engineering, for living organisms dynamic stability by feedback is often called ‘
homeostasis’, following W. B. Cannon's pioneering book
The wisdom of the body (1932). In the history of engineering, there are hints of the principle back to ancient Greek devices, such as self-regulating oil lamps. From the Middle Ages the tail vane of windmills, continuously steering the sails into the veering wind, are well-known early examples of guidance by feedback. A more sophisticated system reduced the weight of the upper grinding stone when the wind fell, to keep the mill operating optimally in changing conditions. Servo-systems using feedback can make machines remarkably life-like. The first feedback device to be mathematically described was the rotary governor, used by James Watt to keep the rate of steam engines constant with varying loads.
Servo-systems suffer characteristic oscillations when the output overshoots the target, as occurs when the feedback is too slow or too weak to correct the output. Changing the ‘loop gain’ (i.e. the magnitude of correction resulting from a particular feedback signal) increases tremor for machines and organisms. It is tempting to believe that ‘intention tremor’ of patients who have suffered damage to the
cerebellum is caused by a change in the characteristics of servo control.
Dynamic control requires the transmission of information. Concepts of information are included in cybernetics, especially following Claud Shannon's important mathematical analysis in 1949. It does not, however, cover digital computing. Cybernetic systems are usually analogue, and computing is described with very different concepts. Early Artificial Intelligence (AI) was analogue-based (reaching mental goals by correcting abstract errors) and there has recently been a return to analogue computing systems, with self-organizing ‘neural nets’.
A principal pioneer of cybernetic concepts of brain function was the Cambridge psychologist Kenneth Craik, who described thinking in terms of physical models analogous to physiological processes. Craik pointed to engineering examples, such as Kelvin's tide predictor, which predicted tides with a system of pulleys and levers. The essential cybernetic philosophy of neurophysiology is that the brain functions by such principles as feedback and information, represented by electro-chemical, physical activity in the nervous system. It is assumed that this creates mind: so, in principle, and no doubt in practice, machines can be fully mindfu.
Richard L. Gregory
Bibliography
Cannon, W. B. (1932). The wisdom of the body. New York.
Craik, K. J. W. (1943). The nature of explanation. Cambridge.
Mayr, O. (1970). The origins of feedback control. Cambridge, M.A.
Shannon, C. E. and and Weaver, W. (1949). The mathematical theory of information. Urbana.
Weiner, N. (1948). Cybernetics. New York.
See also
balance;
homeostasis;
proprioception;
vestibular system.
